John Martin Fischer and Compatibilism

So the next view in “Four Views on Free Will” is John Martin Fischer’s view on Compatibilism, which includes his unique view of compatibilism called semi-compatibilism.  What I’m going to do here is talk a bit about why compatibilism is attractive (since he does that at the beginning of the chapter) and then talk a bit about semi-compatibilism since I don’t think it really achieves what he wants it to achieve.

First, Fischer focuses on arguing that it seems obvious that we make choices, but also notes that determinism could be true, so a compatibilism that allows us to retain decision-making abilities even if it turns out that determinism is true.  Unfortunately, this way of talking about it seems to bias him towards simply making it possible for us to retain “free will” if determinism happens to turn out to be true, which isn’t going to satisfy either of the other two main sides — yes, there is a third one lurking out there since it’s the last segment of the book — because hard determinists think that determinism is true and libertarians are going to reply that there definitely seem to be types of determinisms that would eliminate all free will, and so finding that there might be types of determinisms that wouldn’t isn’t all that impressive, especially since the hard determinists are talking about determinisms that really do seem like they would eliminate all free will.  So compatibilists need to do more than simply carve out an exception in some determinisms for free will, but instead need to show why the determinisms we are likely to have don’t cause any issues for free will.  And I think that this impacts his semi-compatibilism, which I’ll get to in a moment.

Before that, though, I’d like to note that he is somewhat right about the appeal of compatibilism, with it being a view that bridges the two positions.  But I think that the main reason it is appealing to many — although many find it implausible rather than plausible — is a basic philosophical methodology.  Our intuitions strongly support the idea that we have some sort of meaningful free will and that we actually make meaningful choices.  However, a lot of the intuitions from science and even every day life suggest that for everything else things are pretty much entirely determined causally (we can ignore quantum phenomena because they don’t seem to have much if any impact at the macro level and we don’t observe quantum phenomena in our everyday lives).  So it raises the obvious question that if everything else in our world is determined, why would we be any different.  So what we have are two strong and pretty plausible and intuitive arguments for either side.  And when we have such a clash of arguments where we can make strong arguments for either side — think Kant’s Antimonies — and can’t see where either side’s argument goes wrong there is a natural philosophical tendency to ask whether maybe the problem isn’t in the arguments for each side, but instead in the arguments that we are making to say that they can’t both be true.  Maybe we’re wrong about that and they can both be true.  And all compatibilisms at their heart say exactly that:  both of those intuitive and strong arguments are, indeed, actually true.

Which means, of course, that the debates against compatibilism really should focus on whether it is indeed the case that both can be true, and so on whether we can have a meaningful free will even if things are totally determined.

Which leads to his semi-compatibilism.  The problem with it is that as I understand it he relies on a distinction between regulative control and guidance control, and argues that while determinism would kill regulative control (which requires access to alternative possibilities that determinism would eliminate) from guidance control which he argues would not or would not need to be impacted.  But those terms can be a bit confusing, so I’d like to focus on his example of the difference (from page 58) which is a Frankfurt-type example (I’ll be summarizing it as I understand it and not quoting it because I have an actual book and that would be a real pain for me):

Imagine that someone is going to vote, but hasn’t decided who to vote for yet.  Unbeknownst to them — but knownst to us — someone else has implanted a chip in their brain that will notify them if they decide to vote for Party B so that they can instantaneously flip a switch and change the decision so that they will instead vote for Party A.  So if they go through their process and decide to vote for Party A, then they would clearly have made a free choice because the chip in their brain is irrelevant to that process.  However, it’s also clear that they were going to vote for Party A no matter what happened.  So the fact that there was only one possible outcome doesn’t mean in this case that they had no free will.  I think Fischer at least analogizes regulative and guidance control to these two examples, making the case where their deliberation process chooses Party A the guidance case and so in a case where we have only that sort of control it looks like we would still be making a free choice even if we really “couldn’t choose otherwise”, since no matter what happened the person was always only going to choose to vote for Party A.

The problem I see here is that I don’t think that any real kind of determinism is going to allow for even that kind of control, and so semi-compatibilism can’t get off the ground.  Let’s alter the example to add that someone else inserts another chip that completely controls the person’s decision-making process to walk them through a decision-making path that ends with them deciding to vote for Party B.  They walk through that process and decide to vote for Party B, and then the chip is triggered and they end up having that decision changed to vote for Party A.  So they didn’t freely choose to vote for Party A at all, I think we can all agree (even Fischer).  However, let’s then change it so that the person who inserted the second chip instead sets it so that their decision-making process gets them to vote for Party A.  As before, the first chip is irrelevant to the outcome, but now we have to ask if the person made a free choice to vote for Party A.  And it looks like they didn’t, because their entire decision-making process was determined by an agent outside of themselves to without fail come to that specific conclusion.  It’s not only the case that they couldn’t have done otherwise, but also that they couldn’t have decided otherwise.  And if they couldn’t even have decided otherwise, there seems no room for any kind of control, whether regulative, guidance, or anything else.

What this shows is that it isn’t sufficient to show that a decision was the result of a deliberation and our decision-making processes to show that we should think that it’s free.  It’s entirely possible that under determinism our entire decision-making process was determined before it was even engaged, and so that it’s just “going through the motions” to produce a predetermined decision, and so has no causal impact on the outcome at all.  And if nothing in our decision making processes has a real and meaningful causal impact on the outcome, then our decisions don’t seem to be free at all, not matter what type of control we talk about.

So Fischer would need, at a minimum, to show how we can have a specific type of control that still remains even if determinism is true.  And he would need to address it in the forms of determinism that are more likely to be true, and hard determinists and libertarians both agree that that’s the very strict causal sense that would determine the entirety of our decision-making processes and so would look a lot like the example of the chip that I added to his Frankfurt example.  I did not see how he could escape that case in his chapter, and so see his semi-compatbilism as having as serious issues as other compatibilisms and so that it doesn’t really give us any advantage here.  It’s true that he can concede that we would lose one type of control and so not have to argue over that anymore, but he doesn’t seem to have shown that the same mechanism that eliminates regulative control doesn’t also eliminate guidance control and every other sort of control that we could have as well.

Next is the position that I have the least sympathy for, which is hard determinism or, as they — and many others, I suppose — call it, hard incompatibilism, which I don’t like as a term because libertarians are hard incompatibilists as well, which is precisely why they are libertarians.

Tags: free will